suggested fix for score-based tests on OpenMx models without labels

[brandmaier]

The following code does not work with current semtree and score-based tests because the model has no labelled parameters. Thus, the score computation returns only zeros and the tree will find no split. I suggest that we identify free parameters not by whether their labels match in the OpenMx matrices but whether the free matrix has respective entries of true or false.

library(OpenMx)
library(semtree)

set.seed(2349)

manifests<-c("DeltaPA")
latents<-c()
model <- mxModel("Simple Model", 
                 type="RAM",
                 manifestVars = manifests,
                 latentVars = latents,
                 mxPath(from="one",to=manifests,
                        free=c(TRUE), 
                        value=c(1.0) , arrows=1#, 
                       # label=c("mu") 
                        ),
                 mxPath(from=manifests,to=manifests,
                        free=c(TRUE), value=c(1.0) , 
                        arrows=2#, 
                       # label=c("sigma2") 
                        ),
                 mxData(tree.data, type = "raw")
);
fitted <- mxRun(model)
ctrl = semtree.control(
  method="score", 
  bonferroni = TRUE)
tree = semtree( model = model, 
                data = tree.data, 
                control=ctrl)